Wind Tunnel White Paper

Our readers have been asking for this article for a LONG time. In August 2013 and February 2014, TriRig made two separate trips to the FASTER wind tunnel, to learn how the Alpha compared to other bars on the market. We knew it was a very fast bar, we just didn't know how fast. And although the results of the first test were very favorable, I had a problem with some of the protocol - that is, I wasn't convinced the test was perfectly fair. So we made the second trip, to perfect the protocol and deliver results that would withstand the scrutiny our sophisticated readers are capable of. Those two trips, taken together, are the basis for this article. While the data is all based on the latter trip, both trips helped to inform our testing, protocol, and provide additional perspective.

Unlike the wind tunnel test of our Omega brake, I will not be presenting a separate "white paper" document for the Alpha. This article is the Alpha white paper, just not in the usual PDF form.

Why? Well, for the Omega, we had the benefit of an independent author, Dr. Andrew Coggan, to write the white paper for us. In that sense, it was a separate, scholarly source. This time, I'm the sole author. So there's no reason to create two publications, since both would contain the same information, results, and biases. However, there's still the additional source of the FASTER tunnel and staff, whose reputation for straightforward and transparent testing carries significant weight throughout the industry. FASTER has been used by players across the cycling and triathlon worlds, from frame and component manufacturers, to trade publications, to individual athletes looking for an edge. I'm proud to have had the opportunity to test the Alpha there, and I hope to do their work justice in the way that I report it.

So, without further ado, let's get to it.

Bars Tested
We tested six different front end setups at FASTER. These included the TriRig Alpha, the original 3T Ventus, the PRO Missile Evo, the Felt Bayonet 3, the Zipp Vuka Alumina, and finally a set of traditional road drop bars with clip-on extensions (taken from the Felt Bayonet 3). While this is not an exhaustive list of the bars on the market, we felt it was fairly representative of the available spectrum of bars.

Testing Protocol
In order to get a legitimate comparison between different aerobars, we attempted to keep as much of the setup identical from test to test, as well as eliminate any unnecessary test equipment that could confound the results. The basic setup was as follows:

Bike: Specialized Shiv frame, TriRig Omega SV brakes, Dash Cycles saddle + seatpost combo. The crank, chain, and derailleurs were removed and the BB shell covered, to avoid any noise in the data caused by slight movements in these parts. Also, brake and shift cables were not present in the test for the same reason. However, some bars have better cable management than others, and those differences are not present in our final data.
Wheels: Mavic CXR 80. The tunnel speed was set at 30mph (the de-facto industry standard), and the wheels were spun up to the same speed. There's a very interesting and little-known fact to mention here. Historically, some tunnels did not have the capability to spin wheels at the same speed as the tunnel (for example, the wheels could only spin at 24mph while the tunnel wind blew at 30mph). FASTER was built from the ground-up specifically for cycling, and wanted to ensure that the wheel speed could always match the apparent wind speed, for the most accurate results possible. Some testers still use mismatched speeds here, if only for more relevant comparisons to their own historical data. However, I was interested in simply getting accurate data, so we spun the wheels at the same speed as the tunnel.
Stem: TriRig Sigma Flat. This stem was used on the TriRig Alpha, Felt Bayonet 3, and Zipp Vuka Alumina bars. The 3T Ventus has an integrated stem. For the Shimano PRO Missile Evo bars, we used the matching PRO Missile Evo stem, as these two were designed to be used together. We first tested the Missile with the Sigma Flat, and found that the drag curve was virtually identical with either stem. This may be because although the Missile stem presents a slightly larger frontal area, it was specifically designed to work with the overall shape of the Missile bar.
Bars: The bars were set up, as closely as possible, with the same pad stack, and where possible, pad stance width. We measured each bar twice: once by me during bar prep, and again by FASTER staff once each bar was actually mounted on the test bike. Pad stack was measured at the trough of each pad (where your arm actually sits). Each setup measured within 5mm of the baseline (a 14.5cm drop from saddle to pad trough). The brake lever holes were plugged with identical hemispherical bar plugs, except for the 3T Ventus, which has (very tiny) non-removeable brake lever blades attached. In my opinion, those Ventus levers are so small as not to contribute enough drag to bias the results.
Extensions: In order to get the most meaningful results, we wanted to use the same extensions across all bars. However, one of the bars, the PRO Missile Evo, has integrated extensions with a proprietary bend and an inward canting, which cannot be swapped for standard bars. In order to normalize them, we truncated the Missile extensions just before their bend, effectively turning them into straight extensions with no canting. For every other bar, we then used straight extensions of the exact same length (21.5cm). That length is slightly shorter than what the average athlete might use in practice, but this was the best way for us to make the results consistent across all bars, and we don't believe the shorter length biased the results. The ends of every extension were plugged with identical hemispherical bar plugs (except for the Missile extensions, which were fitted with their own proprietary plugs to fit their oval cross sections).
Tunnel Shreds: For each wind tunnel sweep (called a "shred"), the wind and wheels spun up to 30mph with the bike at 0 degrees (pointing straight into the tunnel). After allowing the wind to settle for approximately 30 seconds, the bike was rotated to -20 degrees of yaw (drive-side towards the wind). At each point, the wind was allowed to settle for an additional 30 seconds, and then 30 seconds of data was taken and averaged, to generate the single data point for that yaw angle. After that, the bike was rotated 2.5 degrees in the positive direction, and the testing cycle repeated, all the way through 0 degrees of yaw and continuing through to positive 20 degrees (non-drive-side towards the wind). This is on the more extensive side as far as testing protocols go. Other tests often use shorter settling times, shorter sampling times, wider intervals between yaw angles, and/or shorter ranges of yaw angle. All of these can help shorten the total length of the test, and/or allow more parts to be tested in the same amount of time. However, the parts being tested here, aerobars, are relatively small and light on drag, so a robust protocol was important to getting meaningful results.